1. Field of the Invention
The present invention relates to an image processing apparatus, an image processing system, and an image processing method for adjusting a recording position of a recording head which discharges ink.
2. Description of the Related Art
In a conventional inkjet recording apparatus, a recording positional deviation occurs depending on accuracy of a manufacturing process of a printing head (recording head) or mounting accuracy of the printing head. For example, a printing positional deviation may occur between a monochrome ink discharge port chip and a color ink discharge port chip, between ink discharge port arrays, between forward printing and backward printing, or between paths during multipass printing.
In order to solve such a recording positional deviation, for example, a plurality of patterns is printed while shifting a printing position under a print condition corresponding to an adjustment item to output a chart. The output chart is read by a scanner, and a pattern position in image data is detected. Luminance of a pattern in a pattern array corresponding to the adjustment item is measured based on the detected pattern position. It is determined that a pattern with the lowest luminance among the luminance of the measured patterns is a pattern of an optimal recording position, and an adjustment value is determined based on a shifting amount of the pattern. Accordingly, printing position adjustment can be realized.
Thus, processing for adjusting the printing position based on the pattern luminance in the image data necessitates pattern position detection. However, in processing for the pattern position detection, there occurs a problem that the pattern position may deviate in the image data due to a deviation of a chart installing position in a reading apparatus, optical distortion of the reading apparatus, or accuracy of a manufacturing process of the reading apparatus.
Under such circumstances, various methods have been discussed to accurately detect a position of a predetermined pattern or object from image data. For example, Japanese Patent Application Laid-Open No. 11-340115 discusses a technique which extracts a predetermined feature amount from a detected image, generates an abstract pattern of the detected image based on the feature amount, and extracts a feature amount using a reference image. An abstract pattern of the reference image is generated based on the feature amount, and a pattern position of the reference image is detected in the image data by matching the abstract patterns with each other.
However, the above described conventional technique has the following problems. For example, in the technique discussed in Japanese Patent Application Laid-Open No. 11-340115, processing is performed to obtain feature amounts such as a density of graphics, a center of gravity, a principal axis of inertia, and moment of the principal axis of inertia from the detected image. Abstract patterns in which the feature amounts are disposed based on the detected image and the reference image are created. The positions of the patterns are detected by matching the created abstract patterns with each other. However, calculation time is necessary for creating the abstract patterns based on the two images, i.e., the detected image and the reference image, so that processing time for pattern position detection may be increased.